In this piece, I take a close look at why the public doesn’t have an accurate understanding of machine learning/AI, and explain what the negative results will be if data science doesn’t work to fix this problem.
Keywords: public policy, education, machine learning
A discussion of the different forms a data science job can take, including practical explanation of what the responsibilities look like. I also discuss the overloading of individual roles and how that can cause burnout.
Keywords: data science, careers
My thoughts on the evolution of job titles and career paths in the DS/ML field, including some discussion of the possible DEI implications.
Keywords: dei, machine learning
My step by step description of how I designed and implemented the Smart Geofences feature at project44.
Keywords: gis, geofences, machine learning, devops, airflow, python
In this article I described my process for starting a new ML project, and gave tips from my experience for newer data scientists.
Keywords: machine learning, tutorial, advice
The hosts of this show gave me an hour to talk about all kinds of issues in DS/ML, including ethics, data privacy, risk/rewards, D&I, and more. I mentioned a couple of things in the episode that are linked below.
Keywords: ethics, machine learning
In this series, I took a close look at these 5 areas of the data science or machine learning development workflow, telling you what the cloud approach looks like versus the local approach, explaining the pros and cons of each, and describing a few tools you might want to try.
Keywords: python, machine learning, cloud computing
Data science model deployment can sound intimidating if you have never had a chance to try it in a safe space. In this blog post, I give the end-to-end explanation of how you go from zero to a deployment that you can share with others.
Keywords: python, HTML, machine learning, deployment, API
List comprehensions are incredibly powerful, but also can be unintuitive and confusing. I wrote a guide to understanding and using them, designed for those who use for-loops in their regular programming.
Keywords: python, programming, beginners
Generative Adversarial Networks (GANs) are an increasingly popular and very powerful form of computer vision deep learning, and they require a whole lot of compute to be done effectively and speedily. This is just the sort of use case where parallelization with Dask clusters can make a difference in your workflow, so I wrote a tutorial on how to do it!
Keywords: python, machine learning, deep learning, dask
I wrote a blogpost to accompany several talks and a github repo all about using different Python visualization tools. Get the code at https://github.com/skirmer/new-py-dataviz.
Keywords: python, data visualization
Do you love pandas, but hate when you reach the limits of your memory or compute resources? Dask gives you the chance to use the pandas API with distributed data and computing. In this article, you’ll learn how it really works, how to use it yourself, and when/if to switch.
Keywords: python, pandas, dask
Lazy evaluation is the core of parallelization, but it doesn’t have to be confusing or complicated — in this guide, learn the basic concepts you need to get started! I use dask to demonstrate but this is useful for anyone trying to get the hang of parallel computation. Keywords: python, parallelization, dask
You should not require or ask for a generic STEM Ph.D. for data scientist candidates. In this blog post, I give a detailed argument on this and discuss the critiques of the practice.
Keywords: data science, social science
I use the Stanford Dogs dataset again, this time to demonstrate accelerating transfer learning to improve Resnet50. The inner workings of how PyTorch supports multi-machine, multi-GPU training can be confusing but I have deciphered it for you here.
Keywords: machine learning, python, deep learning
I use the Stanford Dogs dataset to demonstrate accelerating an image classification problem with GPU Clusters. If you have been thinking about GPUs but don’t know where to start, or what they might be good for, I recommend this as a place to start!
Keywords: machine learning, python, deep learning
Job scheduling is what takes academic machine learning to production level for real business or project value. I went through three tools (cron, Airflow, and Prefect) in this article and discussed the pros and cons to each. Depending on your task and circumstances, any one of these tools might be what you need.
Keywords: python, job scheduling, airflow
In concert with my presentation at ODSC Europe 2020 I wrote a blog post to discuss why and how you might use scheduled jobs to make your data infrastructure work better for modeling and machine learning.
Keywords: job scheduling, airflow, data warehousing
stephanie@stephaniekirmer.com
kaggle
|
github
|
linkedin
|
youtube